Inpainting
Editing or filling in parts of an image that are already there, or extending images beyond their original borders.
Inpainting is a way to change or add to an image. It can fill in missing spots, like if a part of a photo is scratched out. It can also be used to add new things to an image based on what's already there.
For example, you could use inpainting to remove an unwanted person from a photo. The AI looks at the pixels around the empty space and generates new pixels to make it look like the person was never there. OpenAI's DALL-E uses a related technique called outpainting to extend images, adding content beyond the original frame.
Inpainting is a generative AI task focused on reconstructing missing or corrupted regions within an image. It leverages surrounding pixel information and learned image priors to synthesize plausible content that seamlessly blends with the existing image. This is typically achieved using diffusion models or generative adversarial networks (GANs) trained on large image datasets. Tradeoffs involve potential for artifact generation or unrealistic content if the context is ambiguous or the model is not sufficiently trained.
Inpainting is a core capability in modern image generation models. Recent advancements in diffusion models have made inpainting more powerful and accessible, allowing for sophisticated image editing and content creation. Tools like OpenAI's DALL-E demonstrate its application in extending image boundaries, a form of inpainting.
- DALL·E: Introducing outpaintingOpenAI · OpenAI News ·
- OpenAI ChatGPT
- OpenAI's DALL-E offers outpainting, a related feature that extends images by generating new content beyond the original borders, effectively using inpainting principles.
- Outpainting
- Outpainting extends an image by adding content outside its original
- Image Generation
- Image generation creates entirely new images from text prompts
- Image Segmentation
- Image segmentation identifies and outlines objects within an
The concept of inpainting originated in the field of image restoration and computer vision, with early research focusing on reconstructing damaged photographs. The term and its application in AI have evolved significantly with the advent of deep learning. Source ↗
Use OpenAI's DALL-E 3 image editor. Upload an image and select an area to erase. Then, provide a text prompt describing what you want to fill that area with. The AI will generate content to replace the erased section, demonstrating inpainting.
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- PromptRole-play a student struggling with math for AI tutor simulationSimulate a student's math struggles and uncertainty for AI tutor evaluation. · 5 min
- TemplateReusable workflow templates for healthcare teamsTeams can create and share templates for common workflows like documentation or prior authorizations. · 15 min
- TemplateTemplate for building an AI tool tailored to your workA structured approach to defining and building a custom AI tool for specific business needs. · 15 min
- TemplateEvals CSV prompt list for agent skillsA CSV file template for listing prompts used in evaluating agent skills. · 20 min
- HeuristicDouble coding agent task completion with docs linksProviding coding agents with a documentation link significantly increases their task completion rate. · 2 min
Questions people ask
- What is the difference between inpainting and outpainting?
- Inpainting fills in missing or masked areas within an existing image. Outpainting extends an image by generating new content beyond its original boundaries.
- Can inpainting create realistic results?
- Modern inpainting models, especially those based on diffusion, can produce highly realistic results by analyzing context and learning image priors. However, results can vary depending on the complexity of the edit and the quality of the model.
First seen on Subvolts Aug 31, 2022. Updated Sep 15, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.